Red Hat introduces the Ansible development tools MCP server, which connects AI assistants like GitHub Copilot, Claude Code, Gemini CLI, Cursor, and Windsurf directly to Ansible CLI tools (ansible-lint, ansible-creator, ansible-navigator) through the Model Context Protocol. Instead of an AI merely describing how to run a command, it can execute linting, scaffolding, environment setup, and playbook runs deterministically, using the developer's actual local environment context to avoid hallucinated module names or missing dependencies. The server ships bundled in the Ansible VS Code extension, is available as a container image or via npx, and is currently a Technology Preview. It pairs with an Ansible Automation Platform MCP server (GA in Ansible Automation Platform 2.7) for job management and inventory queries at the operational layer, with RBAC-scoped tokens recommended for non-production use.
Table of contents
The Ansible playbook context-switching problemModel Context Protocol: From advisor to operatorWhat the Ansible development tools MCP server can doBring your own modelDeveloping an Ansible playbook with a deterministic coding assistantWhat the Ansible MCP server looks like in practiceGetting started with the Ansible MCP serverBeyond development: Using the Ansible MCP server in execution environmentsAdditional resourcesQuestions this post answers
How do I connect Claude Code to the Ansible development tools MCP server?
Run 'claude mcp add ansible -- npx -y @ansible/ansible-mcp-server --stdio' in the terminal. This registers the Ansible development tools MCP server, giving Claude Code access to tools like ansible-lint, ansible-creator, and ansible-navigator with structured, deterministic calls instead of generic shell command suggestions. daily.dev surfaces setup details like this for developers wiring AI assistants into their Ansible workflow.
What can the Ansible development tools MCP server do that a regular AI chat assistant can't?
It executes ansible-lint, ansible-creator, and ansible-navigator directly with correct flags and environment awareness, rather than just describing commands to run manually. It grounds suggestions in the developer's actual installed collections and module versions via ansible-doc, preventing hallucinated collection names or nonexistent module parameters, and can auto-fix lint violations until the code is clean. Developers weighing AI-assisted automation tooling track capability breakdowns like this on daily.dev.
Which Ansible Automation Platform version includes general availability of the MCP endpoints on the gateway?
MCP endpoints for job management, inventory queries, and system monitoring are available starting with Ansible Automation Platform 2.5 and reach general availability in Ansible Automation Platform 2.7. Platform engineers can use these endpoints with role-based access control, scoping tokens as read-only for queries or read-write when job launching is required. Platform engineers tracking Ansible Automation Platform release milestones follow updates like this on daily.dev.